0% found this document useful (0 votes)
9 views33 pages

MAS632 - Lecture 1

This document provides an introduction to a course on management science models for decision making. It discusses the relationship between management science and operations research (O.R.), noting that the course will focus on O.R. It also defines analytics and its three categories. Additionally, it discusses decisions and models, explaining that models are simplified representations used to make better decisions. The document outlines some benefits of modeling and provides examples of successful applications of O.R. It also introduces optimization problems, linear programming problems, and provides an example farmer's decision problem formulated as a linear program.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
9 views33 pages

MAS632 - Lecture 1

This document provides an introduction to a course on management science models for decision making. It discusses the relationship between management science and operations research (O.R.), noting that the course will focus on O.R. It also defines analytics and its three categories. Additionally, it discusses decisions and models, explaining that models are simplified representations used to make better decisions. The document outlines some benefits of modeling and provides examples of successful applications of O.R. It also introduces optimization problems, linear programming problems, and provides an example farmer's decision problem formulated as a linear program.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 33

SCHOOL OF BUSINESS ADMINISTRATION

MANAGEMENT SCIENCE DEPARTMENT

Welcome to MAS 632

Management Science Models for


Decision Making

1
Management Science vs. O.R.
Management Science is a very broad term that
includes many aspects of the science of managing
a business

Operations Research (O.R.) is a more precise


name to define “the discipline of applying
advanced analytical methods to help people make
better decisions” a.k.a. “the science of better”

This course is about O.R.

INFORMS is the main O.R. society in the U.S.A.


See www.informs.org and www.scienceofbetter.org
2
Have You Heard of “Analytics”?
It’s one of the latest buzzwords!

Refers to the application of statistics, computer


science, and O.R. techniques to make better use
of data

Divided into three categories:


Descriptive Analytics:
Look at data to understand what has happened (e.g. regression,
statistics)
Predictive Analytics:
Try to predict what could happen in the future (e.g. forecasting,
data mining, statistics)
Prescriptive Analytics:
Find best course of action to increase chances of realizing the best
outcome (e.g. simulation, optimization)
3
Decisions and Models
In business (and life) we make a number of
important decisions everyday
To make better decisions we use models

What is a model?
A simplified representation of a real-life object,
phenomenon, situation, etc.
Ex: a map, a miniature, a set of math formulas

How simplified should the model be?


Too simple à useless (too far from reality)
Too complex à useless (intractable)
Only keep what is relevant (may still be too complex)

4
Benefits of Modeling
Insight and understanding
Models give us insight into complicated problems
thereby improving decision making

Economy
It is less costly to analyze decisions using models
than in a real-world experiment

Timeliness
Models deliver needed information more quickly
than their real-world counterparts

Feasibility
Models can be used to test extreme conditions or
situations that rarely occur in the real world
5
Why Should You Care?

Operations Research is everywhere around you and,


yet, most of you didn’t even know it existed

Most successful business these days use O.R. in one


form or another. Those that don’t would probably
benefit a great deal from it

HBR classified O.R. as one of the Top Ten Most


Influential Management Ideas of the Millennium

Let’s go over a few examples…

6
Some Success Stories
Created a vehicle routing and scheduling system to
run its delivery and home service fleets more
efficiently. Savings: $42M per year

Redesigned its overnight delivery network.


Savings: $87M in 2 years

Improved their workforce scheduling and reduced


operational costs. Savings: $150M per year

Optimized the way it designs and tests vehicle


prototypes. Savings: $250M

Used O.R. techniques to reduce inventory.


Savings: $1 billion!
7
Some Other Applications of O.R.
Finance: portfolio selection

Marketing: online advertising, allocation of TV ad slots

Accounting/Auditing: budget allocation for the IRS

Auctions: winning bidder selection

Health Care: cancer treatment, organ allocation

Homeland Security: criminology, passenger


screening at airports, disaster recovery planning

Many, many more… I’ll show you more examples


throughout the course
8
Our Kind of Model

O.R. deals with mathematical models

Why use a mathematical model?


Can be implemented in a computer
Model output can be analyzed
Spreadsheet models are examples of
mathematical models used by managers today

9
A Sample of Mathematical Models

Profit = Revenue – Expenses

Y = log(X1 + X2 * X3)

10 5
z = åå aij xij
i =1 j =1

10
The Problem Solving Process

Identify Formulate &


Problem Implement Analyze Test Implement
Model Model Results Solution

unsatisfactory
results

11
Main Messages
A decision model is a tool, not a substitute for the
human decision maker

The results (or outputs) of a model should always


be analyzed in a critical way

A structured approach to decision making can help


us make good decisions, but cannot guarantee
good outcomes

There are plenty of success stories showing the


benefits of analytical methods in the business
environment; see optional readings on Blackboard!

12
SCHOOL OF BUSINESS ADMINISTRATION
MANAGEMENT SCIENCE DEPARTMENT

Introduction to Optimization
and Linear Programming

13
Introduction to Optimization
We all face decisions about how to use or
allocate limited resources such as:
Oil in the earth
Land for dumps
Time
Money
Workers

Due to resource scarcity, we often seek


the most efficient way of using them to
achieve our objectives.
14
Mathematical Programming (MP)
MP is a field of Management Science (or O.R.) that
studies techniques to find the optimal, or most efficient,
way of using limited resources to achieve the objectives
of an individual or a business
a.k.a. Optimization

Applications of optimization:
Determining Product Mix
Manufacturing and SCM
Transportation, Routing and Logistics
Portfolio design
Sports scheduling
Etc.

15
Optimization Problems
Characteristics of optimization problems
Decisions
Variables under management (your) control that
can impact the performance metric
Constraints
Imposed by resource limits
Imposed by the structure of the problem
Objectives
Goal of the optimization problem
Maximize or Minimize performance metric

16
General Form of Optimization Problem
Max (or Min): f0(X1, X2, …, Xn) Objective
Subject to: f1(X1, X2, …, Xn)<=b1
: Decision
fk(X1, X2, …, Xn)>=bk Variables
:
fm(X1, X2, …, Xn)=bm Constraints

Note: If all the fi in an optimization problem are linear,


the problem is a Linear Programming (LP) problem

17
Linear Programming (LP) Problems
Max (or Min): c1X1 + c2X2 + … + cnXn
Subject to: a11X1 + a12X2 + … + a1nXn <= b1
:
ak1X1 + ak2X2 + … + aknXn >=bk
:
am1X1 + am2X2 + … + amnXn = bm

Note: Every LP can be rewritten to use a max


(or min) objective and only one type of
constraint (either =, <= or >=)

18
Example: A Farmer’s decision problem
A farmer owns 60 acres of land where she may
plant wheat and corn
An acre planted with wheat yields $200 profit
An acre planted with corn yields $300
Planted acres use fertilizer and labor as follows
Wheat Corn
Labor (workers) 3 2
Fertilizer (tons) 2 4

There are 100 workers and 120 tons of fertilizer


available

Q: How can she maximize total profit?


19
5 Steps In Formulating LP Models
1. Understand the problem
2. Identify the decision variables
3. State the objective function as a linear combination of
the decision variables
4. State the constraints as linear combinations of the
decision variables.
5. Identify any upper or lower bounds on the decision
variables

Hint: If you get stuck in steps 3 or 4, you may need to


rethink your choice of variables

20
The Farmer’s Problem
W = # acres to plant with wheat
C = # acres to plant with corn

Maximize 200W + 300C


Subject to
W + C ≤ 60
3W + 2C ≤ 100
2W + 4C ≤ 120
W ≥ 0, C ≥ 0

21
22
Solving LP Problems: Graphical Approach

The constraints of an LP problem


define its feasible region

The best point in the feasible region is


the optimal solution to the problem

For LP problems with 2 variables (i.e.


in 2 dimensions) it is easy to plot the
feasible region and find the optimal
solution
Can also do it in 3D, but it’s tedious…
23
Summary: Graphical Solution to LP Problem

1. Plot the boundary line of each constraint


2. Identify the feasible region
3. Locate the optimal solution by either:
a. Plotting level curves (see book 2.10.5-6)
b. Enumerating the corner points

Note: Corner points are also called extreme points


or vertices

24
Special Conditions in LP Models

A number of anomalies can occur in LP


problems
Alternative Optimal Solutions
Redundant Constraints
Unbounded Solutions
Infeasibility

25
Ex: Alternative Optimal Solutions
X2

250

objective function level curve


200

150

100

alternative optimal solutions


50

0
0 50 100 150 200 250 X1
26
Ex: Redundant Constraint
X2

250
boundary line of constraint 1

200
boundary line of constraint 3

150

100 boundary line of constraint 2

Feasible Region
50

0
0 50 100 150 200 250 X1
27
Ex: Unbounded Solution (max problem)
X2

1000 objective function


X1 + X2 = 600 -X1 + 2X2 = 400

800
objective function
X1 + X2 = 800

600

400

200

X1 + X2 = 400
0
0 200 400 600 800 1000 X1
28
Ex: Infeasibility
X2

250

200

feasible region for


150 constraint
X1 + X2 >= 200

100

feasible region
50 for constraint

X1 + X2 <= 150
0
0 50 100 150 200 250 X1
29
The Diet Problem
Suppose we have several foods, each with
nutritional content as follows

food A C B1 $/serving
beef 60 20 40 4.00
chicken 80 0 20 2.40
spaghetti 25 40 25 3.00
turkey 60 20 15 2.50

Minimum required weekly intake of each nutrient is 700

Find least expensive combination of foods that


satisfies the nutritional requirements of a person
30
Modeling the Diet Problem
B, C, S and T indicate how many servings of beef,
chicken, spaghetti and turkey we eat, respectively.
They are the decision variables

Minimize 4B + 2.4C + 3S + 2.5T


Subject to
60B + 80C + 25S + 60T ≥ 700
20B + 40S + 20T ≥ 700
40B + 20C + 25S + 15T ≥ 700
B ≥ 0, C ≥ 0, S ≥ 0, T ≥ 0

31
Another Example
Blue Ridge Hot Tubs produces two types
of hot tubs: Aqua-Spas & Hydro-Luxes.
Aqua-Spa Hydro-Lux
Pumps 1 1
Labor 9 hours 6 hours
Tubing 12 feet 16 feet
Unit Profit $350 $300

There are 200 pumps, 1566 hours of


labor, and 2880 feet of tubing available.
Spend at least 480h producing Hydro-
Luxes
32
Modeling Blue Ridge’s Problem
X1 = number of Aqua-Spas to produce
X2 = number of Hydro-Luxes to produce

Maximize 350X1 + 300X2


Subject to X1 + X2 ≤ 200
9X1 + 6X2 ≤ 1566
12X1 + 16X2 ≤ 2880
6X2 ≥ 480
X1 ≥ 0, X2 ≥ 0

33

You might also like